Preprocessing

This is an R Markdown Notebook. When you execute code within the notebook, the results appear beneath the code.

Try executing this chunk by clicking the Run button within the chunk or by placing your cursor inside it and pressing Cmd+Shift+Enter.

rm(list = ls())
require(readr)

## Loading required package: readr

require(tidyr)

## Loading required package: tidyr

require(dplyr)

## Loading required package: dplyr

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

state_fips <- read.table('data/raw/state_fips.txt', sep = '|', header = T)
county_fips <- read.table('data/raw/county_fips.txt', sep = '|', quote = '', header = T)
county_fips$fips <- 1000 * county_fips$STATEFP + county_fips$COUNTYFP
fips <- left_join(county_fips, state_fips, by = c('STATE', 'STATEFP'))
fips

population <- read.csv('data/raw/voting_eligible_population.csv', col.names = c('County', 'MaleAdultNative', 'MaleAdultNaturalized', 'FemaleAdultNative', 'FemaleAdultNaturalized'))
population[, 2:5] <- apply(population[, 2:5], 2, parse_number)
population$Population <- apply(population[, 2:5], 1, sum)
for (i in rownames(population)) {
  if (is.na(population[i, 'Population'])) {
    population[i, 'Population'] <- population[as.character(as.numeric(i) + 1), 'Population']
  }
}
population <- population %>%
  subset(County != population[2, 'County']) %>%
  separate_wider_delim(County, ', ', names = c('County', 'State'))

#left join (various territories are removed)
population <- left_join(population, fips, by = c('County' = 'COUNTYNAME', 'State' = 'STATE_NAME'))
population

returns <- read.csv('data/raw/countypres_2000-2020.csv') %>%
  subset(year == 2020) %>%
  group_by(county_fips, county_name, state) %>%
  summarize(totalvotes = mean(totalvotes))

## `summarise()` has grouped output by 'county_fips', 'county_name'. You can
## override using the `.groups` argument.

colnames(returns)[1] <- 'fips'
returns

# problems: Alaska, Kalawao County (Hawaii), Puerto Rico (can't vote?)
returns <- full_join(returns, population, by = 'fips')
returns$turnout.rate <- returns$totalvotes / returns$Population
returns

subset(returns, is.na(Population))

subset(returns, is.na(totalvotes))

characteristics <- read.csv('data/raw/cty_covariates.csv')
characteristics$fips <- 1000 * characteristics$state + characteristics$county
characteristics

data <- full_join(characteristics, returns, by = 'fips') %>%
  select(
    State,
    County,
    fips,
    frac_coll_plus2010,
    foreign_share2010,
    med_hhinc2016,
    poor_share2010,
    share_white2010,
    share_black2010,
    share_hisp2010,
    share_asian2010,
    gsmn_math_g3_2013,
    rent_twobed2015,
    singleparent_share2010,
    traveltime15_2010,
    emp2000,
    ln_wage_growth_hs_grad,
    popdensity2010,
    ann_avg_job_growth_2004_2013,
    job_density_2013,
    turnout.rate
  )
data

subset(data, is.na(turnout.rate))

subset(data, is.na(State))

write.csv(data, 'data/processed/data.csv')